Minimizing Farming’s Water Needs
Growing food for humans to eat takes enormous amounts of water — amounts so enormous, in fact, that agriculture accounts for no less than 70 percent of all human uses of water, and even more in developing countries — up to 95 percent in some cases. The average person can get by on a little more than two litres a day for drinking, but it will take somewhere between two thousand and five thousand litres to grow the food that person must eat (the higher number is mostly for meat eaters). That’s about one litre of water per calorie consumed. David Molden, in a monograph for the International Water Management Institute, provides this striking image:
Imagine a canal 10 meters deep, 100 meters wide, and 7.1 million kilometers long — long enough to encircle the globe 180 times. That is the amount of water it takes each year to produce food for today’s 6.5 billion people. Add 2–3 billion people and accommodate their changing diets from cereals to more meat and vegetables and that could add another 5 million kilometers to the channel of water needed to feed the world’s people.1
By extension, enormous numbers of people — and we’re getting more numerous all the time — will multiply this number commensurately. By conservative estimates, it will take an additional 1.3 billion litres of water just to feed the extra mouths that will be born between now and 2025. That, as Sandra Postel, head of the Global Water Policy Project, once calculated in an equally striking image, is the equivalent annual flow of ninety-seven Colorado rivers — and where are we to find ninety-seven new Colorados? Ninety percent of the increased demand will be in developing countries.
It is these numbers that have persuaded economists to stop measuring farm yields in tonnes per hectare; water is now the limiting factor in food production, not land, and so yields are now commonly calculated in kilograms of food produced per tonne of water.
More recently, Postel has neatly reconceptualized agriculture’s water problem:
The search for solutions needs to begin with a reframing of the question. Instead of asking where we can find 97 Colorado Rivers’ worth of water, the question is: How do we provide healthy diets for 8 billion people without going deeper into water debt? Framed this way, the solutions focus on getting more nutritional value per drop of water used in agriculture, which is the key to solving the water-food dilemma. There are many ways we can grow more food for the world with less water, with most falling into four broad categories: (1) Irrigate more efficiently; (2) boost yields on existing farms, especially rain-fed lands; (3) choose healthy, less water-intensive diets; and (4) use trade to make the smartest use of local water.2
I’d add a fifth: develop new salt-tolerant, water-thrifty, and high-yielding crops. And possibly a sixth: turn over management of irrigation water to those who actually use it.
Postel has made another striking calculation: Reducing irrigation water consumption by a mere 10 percent could free up enough water worldwide to meet all the new urban and industrial demands anticipated for 2025.
And 10 percent is an easy target. We can almost certainly do better.
Irrigating More Efficiently
About four-fifths of the world’s croplands depend entirely on rain for water. It’s the last fifth that relies on irrigation, but irrigated lands produce more than 40 percent of the world’s food. Improving irrigation productivity is a good place to start looking for reform.
Until recently, irrigation techniques hadn’t really changed since ancient times — and irrigation is almost as old as farming itself, going back at least five thousand years in central Asia and a great deal further in Mesopotamia. It was used in Imperial China, in old Laos, in ancient Africa, in Tanzania and Zimbabwe, and in America before the Spanish ever laid eyes on the place. For millenniums, farmers have used aqueducts and ditches to take water to where it was needed, and furrows with crude sluices to divert the water to their crops, giving them increased yields and lessening their dependence on the caprices of the weather. But on any real scale, irrigation is not much more than a century old; it needed the massive dams and water diversions of the modern era to make it possible.
The numbers are revealing: two hundred years ago, total irrigated land probably amounted to as little as 6 or 7 million hectares, not much bigger than Long Island or metropolitan Los Angeles and its exurbs. By 1900, the total had jumped to some 50 million hectares, and that number nearly doubled in the half century that followed. Of the estimated 250 million hectares currently under irrigation, half was added in the last three decades, and dozens of countries, ranging from little Israel to massive China, rely on irrigation for much of their domestic food production — Pakistan up to 80 percent. Most of North America relies on rain-fed farming, but without irrigation, agriculture in on the High Plains and in California’s Central Valley, one of the world’s most productive vegetable and fruit resources, would hardly be possible.
In general, farm yields on irrigated land are nearly double the yields found elsewhere. Without irrigation, yields in the world’s major breadbaskets — on which the feeding of the planet is dependent — would drop by almost half. But that irrigation is not without its problems. While it is true that more land is being brought under irrigation, a good deal of formerly irrigated land is now being taken out of production too. Salination of soil to a degree that inhibits farming is spreading at a rate of almost a million hectares a year. In Pakistan alone, two million hectares have been decommissioned, the soil poisoned by high salinity, and farm yields in those area are down 30 percent. Egypt is showing similar declines. In the US Imperial Valley, more land is being decommissioned than commissioned, again because it is overly salty. Salinity and the increasing concentrations of noxious substances in water is the vulnerable underbelly of the irrigation revolution, threatening huge areas of otherwise productive lands. The problem is simple to state, though not easy to fix: without great care and skillful management, large-scale irrigation causes waterlogging, depletion, and pollution of the water supply, and rising salinity in the soil. Left unchecked, these problems can eventually kill the soil altogether.
A summary sheet put out by the US Salinity Laboratory outlines the problems succinctly:
Application of irrigation water results in the addition of soluble salts such as sodium, calcium, magnesium, potassium, sulfate, and chloride dissolved from geologic materials with which the waters have been in contact. Evaporation and transpiration (plant uptake) of irrigation water eventually cause excessive amounts of salts to accumulate in soils unless adequate leaching and drainage are provided. Excessive soil salinity reduces yields by lowering plant stand and growth rate. Also, excess sodium under conditions of low salinity and especially high pH can promote slaking of aggregates, swelling and dispersion of soil clays, degrading soil structure, and impeding water and root penetration. Some trace constituents, such as boron, are directly toxic to plants.3
A second problem with irrigation is water loss that compounds the salination issue. For irrigation, river water is either diverted to canals or directly taken up by pumping. Much of that water is lost in the process. As a working average, almost 60 percent of the water intended for irrigation never gets to the croplands, and a further percentage that does get there never actually gets to the plants that need it. Leaking pipes, unlined canals, evaporation from open reservoirs and canals, and poorly directed spraying cause much of the water to be wasted. Some of this wastage returns to the groundwater, so it is not entirely lost, except in the sense of adding to farmers’ costs.
None of this means that irrigation has been a bad thing or that it is necessarily doomed to failure. There are many ways of increasing water efficiency in agriculture, and all of them have a beneficial effect on salinity. The most obvious way of using water more efficiently is to reduce wastage. In many places where surface water is used for irrigation, it is brought in, sometimes for long distances, via canals or conduits, and it is far from unusual for seepage losses to amount to two-thirds of the water that sets out from reservoirs. Lining canals with plastic sheeting or with concrete can eliminate most of these losses, but it is expensive to do so. When the remaining water does get to the field, it is usually distributed in furrows, which can be diverted by the simple expedient of a shovel from one furrow to the next. Even so, furrow irrigation is wasteful, losing another 10 to 15 percent of the supplied water.
Therefore, more efficient field-distribution systems would be definitely useful. These can be either the centre-pivot system so widely used in the American Great Plains or overhead sprinklers. Both of these lose some water to evaporation, especially in dry areas, and in some jurisdictions farmers are now asked to spray at night rather than during the day to offset this. These systems spread water in the interstices between rows, and so while they are better than furrows, they are still not good. The best system is drip irrigation in which water is delivered through pipes and drip emitters directly to the roots of plants. Famously thirsty and inventive Israel pioneered the technique. It can be done expensively, as high-end California winemakers do it (with miles of buried hoses, the water measured to a fare-thee-well, allowing the judicious admixture of tiny amounts of organic fertilizer), or cheaply, with a bucket or a barrel and a hose. Either way, it can cut irrigation water use in half. According to numbers from Lester Brown of the Earth Policy Institute, Jordanian farmers who used the technique reduced water use an average 35 percent; in India, numerous studies showed gains in water productivity ranging from 46 to 280 percent. Just as California winemakers do, other farmers have begun employing what they call “water stress management,” watering only when necessary and only just enough.4
Another Israeli invention that has similarly increased water efficiency is the use of laser technology to get fields absolutely level, minimizing wasteful runoff and pooling. In fields that are irrigated with flood or furrow systems, this alone can decrease water use by 10 percent or so and increase yields.
One of the initiatives of Sandra Postel’s project, with World Bank help, is to make drip irrigation systems as cheap as possible, and to get them to as many farmers as possible. Traditional systems are relatively expensive, somewhere around $300 per hectare. Bucket and plastic-hose devices are much cheaper, though they are also much more labour-intensive, since they have to be moved from row to row every hour or so. But in countries where labour is abundant and unemployment high, the technology works well. As Lester Brown put it, “These simple systems can pay for themselves in one year. By simultaneously reducing water costs and increasing yields, they can dramatically raise incomes of smallholders.”5
David Bainbridge, a dryland farming specialist at the University of California, Riverside, believes that drip irrigation is not always useful in remote and low-tech parts of the world — the pipes tend to clog and animals eat the tubing. He has experimented with an even lower-tech solution: the buried clay-pot system that he believes is excellent for small-scale farmers and gardeners. Water seeps out through the wall of a buried unglazed clay pot at a rate that is influenced by the plant’s water needs: “When I began the buried clay pot trials, I found that the water efficiency was good, and I kept up with the research. In India, for example, melon yield with the buried clay pot system was 25 tons/hectare using only 2 centimeters of water, compared with yields of 33 tons/hectare using 26 centimeters of water with flood irrigation. But unfortunately scientists don’t typically study these traditional practices.” More recently, he has experimented with what he calls wick irrigation, in which a series of holes is punched into the buried pot and porous wicks inserted, providing a slow and steady source of moisture to plants. “I conducted experiments in the California desert on these systems and found 2-10 times greater efficiency than drip irrigation systems,” he says.6
Another low-tech device that can take the guesswork out of irrigation (and thus uses less water) is something called a tensiometer, typically a sealed water-filled porous cup with a vacuum gauge at the top; as soil moisture decreases, the water level in the tube of the gauge goes down. Some of the newer and simpler tensiometers do away with the gauges and have colour-coded bands instead, making them both easier to read and less prone to fail. In a test for the Columbia Water Project, farmers who used them reported an average of 22 percent water savings and 24 percent energy savings.
Boosting Yields on Rain-Fed Lands
Can rain-fed (non-irrigated) farming improve water use in the same way? After all, 60 percent of the world’s food is still produced with rain as the primary water source. Farming techniques in much of the areas that depend on rain haven’t changed in millenniums — in fact, Rome at the time of Christ was slightly more productive than most farmers in poor countries today. David Molden has calculated that three-quarters of the extra food we will need over the next few decades could be produced just by bringing the productivity of low-yield farms to within 80 percent of high-yield farms. Interestingly, new US figures suggest that although massive agribusiness farms produce more calories per dollar invested, small farms using mixed systems of crops and livestock can actually produce more calories per hectare, done, as Molden suggests, with better water management “and non-miraculous changes in policy and production techniques.” These include improving soil moisture by capturing local rain in small reservoirs to apply to crops when the rains fail through some of the low-cost irrigation systems outlined above. Small but efficient farms in the United States can produce 1.3 kilograms of wheat per tonne of water used; by comparison, in the Punjab area of Pakistan and India, the yield is less than half a kilo.
Sometimes boosting yields can be as simple as managing farming techniques. A system called System of Rice Intensification (SRI) reliably produces higher than normal yields with no fertilizer and no pesticides. It depends on four interlocking ideas, all of which sound too simple but which have been amply proven in trials. First, select and nurture only healthy plants at the start of the growing cycle and transplant them at an earlier age; second, reduce plant numbers — plant single seedlings instead of clumps, giving each plant more room to grow above and below the ground; third, enrich the soil with organic matter, keeping it well aerated; finally, avoid flooding plants and apply water parsimoniously, in ways that favour root and soil-microbial growth.
A young farmer from the state of Bihar, in northern India, Sumant Kumar, came to hold the world record for crop yield on his smallholding, using SRI techniques. He managed to get 22.4 tonnes per hectare (he only has two hectares), whereas even with commercial fertilizers, average yields hardly exceed 8 tonnes — and the average yield worldwide is only 4 tonnes. SRI techniques have since been used in a variety of environments, including the arid north of Mali around Timbuktu. The system is labour-intensive and resistant to large-scale production, but its supporters point out that millions of small farms have more available labour than money, and scaling up is not necessarily in their purview.
Another way of managing farming techniques and to preserve water is to use more under-cropping, agroforestry, and cover crops, and to practise no-till farming.
A simple technique called STRIPS, which stands for, obscurely, Science-based Trials of Rowcrops Integrated with Prairie Strips, has boosted farm yields “spectacularly,” according to farmers who have tried it. All it amounts to is taking around 10 percent of the planted hectarage, usually the least productive part, and replanting it with indigenous prairie plants. The New York Times’s Mark Bittman quotes a plant researcher at Iowa State University, Lisa Schulte Moore: “It’s well-known that perennials provide a broader sweep of ecological function than annuals, so our hypothesis was that if you put a little bit of perennials — a little bit of prairie — in the right place, you get these disproportionate benefits. That is, without taking much land out of production, you get a lot of environmental benefit.” The numbers seem, as Bittman put it, impressive: if you convert 10 percent of a field of row crops to prairie, soil loss can be reduced by up to 95 percent, nutrient loss by 80 to 90 percent, and water runoff by 44 percent. Biodiversity nearly quadruples, and some of those species are pollinators, predators of pests, or both.7
Even simpler is adopting a system of no-till farming: cultivating perennial crops that need no soil cutting. Tillage eliminates weeds but also exposes non-living organic matter to oxygen, releasing their carbon dioxide. David Montgomery, a scientist with the Earth and Space Sciences Department at the University of Washington, in Seattle, has compiled data from around the world that suggest pretty conclusively that conventional agriculture degrades soil much faster than it can be created. Montgomery has concluded that no-till farming could reduce erosion to levels close to soil production rates, and that organic farming methods have been shown to be capable of preserving — and in the case of degraded soils, improving — soil fertility.8 An Australian farmer, Angus Maurice, calls this “no-kill” farming. Wes Jackson, who has worked in the field of sustainable agriculture for more than three decades (one of his creations is the Land Institute, in Salina, Kansas), is another practitioner of no-till farming, though he calls it “natural systems agriculture.” His idea is to develop perennial versions of the major food crops like wheat, sorghum, corn, and rice, crops that have deep roots and don’t have to be replanted each year, thus doing away with the need to till the soil while also radically reducing the need for additions of water. The ancillary benefit is obvious: eliminating nitrogen fertilizers alone would remove two gigatonnes of carbon dioxide a year from the global atmosphere.
Choosing Less Water-Intensive Diets
This is Postel’s third rail, and it risks taking us into nanny-ish territory. It is certainly true that each gram of protein consumed through rice takes five times less water than a gram of beef, or, if you want to measure calories instead of pure protein, it takes twenty times more water per calorie to take beef to table than rice. Two food scientists, Wesley Wallender of University of California, Davis, and Daniel Renault, recently with the Food and Agriculture Organization (FAO), have calculated, surely in jest, that a “balanced diet” consisting entirely of groundnuts, potatoes, onions, and carrots could be produced by using about a thousand litres of water per person per day, compared with the consumption of the steak-eating Americans or Argentinians, who currently need fifty-four hundred litres per day. They suggest that if everyone in the developed world reduced eating animal products by a quarter, it would generate 22 percent of all the additional water requirements expected by 2025. (They don’t explain how water saved in, say, the United States, would benefit anyone anywhere else, since there is no way of transporting that newly liberated water any place it might be needed.)9 Postel herself, who clearly is fond of her Colorado analogy, says that if all Americans cut back on animal products by half, water demand would be reduced by 261 billion cubic metres a year, “a savings equal to the annual flow of 14 Colorado Rivers.”
It must be said that this is the least probable of the reforms. The evidence suggests, in fact, that the trend is the other way, toward more meat eating, as more societies grow wealthier and adopt Western eating habits.
Using Trade to Make the Smartest Use of Water
“Virtual water” is the water embedded in a product — that is, the water used to grow grain or to raise a cow or to make an iPhone. When that product is imported by another country, the embedded water is imported with it: it becomes water the importing country didn’t need to use to acquire the product concerned. This has often been interpreted as a modern form of colonialism, as rich countries once more exploit poor countries by depriving them of a precious resource. In some cases, it is: growing roses in water-scarce Kenya for export to water-rich England may be good for trade but it is horrid for water. But in most cases it is nothing of the kind; it is merely exploiting the differentials in water productivity in different places. If a water-rich country can export embedded water to a water-poor country, it becomes a solution to shortages, an indirect means of transferring water in bulk from one place to another. Israel, for example, has chosen not to grow grain, a water-hungry crop, instead importing it from a country like Canada that has no such shortage. Israel saves between one thousand and three thousand cubic metres of water for every tonne of grain imported. Sandra Postel has estimated that about a quarter of the world’s grain trade is by countries choosing to import water indirectly in the form of grain. As she points out, this can be a good strategy to avoid over-pumping of groundwater or diverting rivers.
There’s another advantage. Virtual water is a way of storing water from good years for use in bad. Except where irrigation is used, all farming areas have poor seasons and good ones, and food storage is a way of smoothing the differences. The numbers are not inconsequential. The FAO’s Daniel Renault calculates that stored grain alone represents a “virtual reservoir” of 500 billion cubic metres of water, or 500 cubic kilometres, a number that rises to 830 billion cubic metres if you add in sugar, meat, and edible oils — this is almost 14 percent of water in existing reservoirs.10 Virtual water is also a way of cheaply desalinating seawater. Somewhere around 8 percent of all virtual water imports is embedded in seafood — which is water that doesn’t come from any country, wet or dry, but represents a net gain of usable water for drylanders.
The only real downside of virtual water is that it risks contributing to regional food insecurity. Countries that come to depend on importing food for their people are taking a chance on prices remaining stable. As countries like Pakistan and China that have large populations and are also water stressed turn to the international grain market, prices will go up. Add this to increasing transportation costs and poor countries risk being shut out from the food market. As Postel put it in her monograph for the Post Carbon Institute, “The food riots that erupted in Haiti, Senegal, Mauritania, and some half dozen other countries as grain prices climbed in 2007 and 2008 are likely a harbinger of what is to come and suggest that a degree of food self-sufficiency may be crucial to food security.”11 Things have eased somewhat since 2009. But in 2014, the world’s food experts were predicting another sharp rise in prices, and that this one would likely stay. A billion people were going hungry, not because there was too little food but because they couldn’t afford to buy it.
Developing New Crops
A fifth way of saving water in agriculture is to grow crops only in those regions to which they are suited, and then, more controversially, to develop new crops suited to arid and saline conditions.
In and around Beijing, to take an obvious example, the regional authorities have banned the growing of rice, a water hog, in favour of more efficient grains. Egypt, a country with no rainfall and only one river, has also banned rice growing, relying instead on imports. In many other parts of the world, cheap and subsidized agricultural water has encouraged the growing of thirsty crops like sugar and cotton in areas without adequate rainfall, such as the American Southwest and the prairies around the Aral Sea. Removing the subsidies make these easy and obvious fixes.
The world also needs to develop new crops entirely, crops that can tolerate the more hostile conditions expected by climatologists or can colonize areas more conventional plants cannot. This is not so much a management issue as it is a science issue, albeit science snarled in politics.
Consider the Green Revolution of the 1970s that ushered in the era of relative abundance with hyper-productive grain varietals, first tested in Mexico and then used, with spectacular success, in Asia. It’s true that the Green Revolution also depended on profligate use of chemical fertilizers and insecticides for its success, but the core was the new grain varieties, developed by food scientist Norman Borlaug, who won the Nobel Prize for his efforts. The Green Revolution saved millions of lives — India and Pakistan particularly would have faced catastrophe without it. It bought us decades of time in which to solve the underlying problems faced by global food production.
Now, we are on the brink of a new agricultural revolution, often called the Blue Revolution.
To some degree, this extends Borlaug’s work, and is a more sophisticated version of what farmers have been doing for millenniums, fine-tuning crops through trial and selection and retrial. An example is the work of scientist Richard Richards, a plant biologist at the Commonwealth Scientific and Industrial Research Organization in Australia. Richards has bred a wheat grain with longer than normal seed sprouts, or coleoptiles, that allow the plant to reach deeper into the soil earlier than conventional seeds. In a series of trials, the new grain increased yields up to 20 percent in arid conditions.12
Another useful example is from Texel, in north Holland, where entrepreneur Marc van Rijsselberghe, together with a plant biologist from Amsterdam’s Free University, Arjen de Vos, has developed a salt-tolerant potato. In 2014, several tonnes of the new potato were on their way to Pakistan to plant in a plot abandoned because of high salinity. There was nothing genetically modified about this new crop, just age-old breeder’s techniques, inspired, as de Vos put it, by studying sea cabbages.13
New and better crops will also come from genetic-modification techniques that aim to introduce beneficial traits from other species and by doing so to increase yields and drought resistance, and boost nutritional values. As I wrote in another context, in my book Our Way Out, any results forthcoming will be controversial:
Few fields are as rife with emotions, some of them bitterly antagonistic and even violent, as GM [genetically modified] foods. Scientists who develop GM crops will bring down a global network of anti-biotech activists on their heads; their labs will be picketed, their field trials sabotaged and their email inboxes flooded with hate mail, actions driven by a toxic mix of emotions, including fear and paranoia, and by “facts” that have only a distant relationship with truth. As a result, many of the GM labs have been obliged to surround themselves with a massive security apparatus, and operate in a world of finely-tuned suspicion.14
These scientists’ position has not been helped by the arrogance, and propensity to lie, of the big agricultural chemical companies that are their patrons. Worse, companies like Monsanto are allowed to patent seed-grain varieties, keeping tight control of their use, and they willingly prosecute farmers into whose fields one of their patented products strays, however innocently. This patenting of life forms has locked farmers into buying annual seeds from the company, and has prevented them replicating their own field grains. To their opponents, such cycles are prima facie evidence of the degeneration of farming into profit. It is all made worse by the often exultant air of some company pronouncements, so when Monsanto’s Robert Fraley is quoted as saying, “We’re now poised to see probably the greatest period of fundamental scientific advance in farming in the history of agriculture,” his opponents hear only the profit-mills grinding.15
It is not hard to find GM successes, even from such as Monsanto. For example, company researchers found a way to mitigate the effect of drought on corn, and in field trials, their new variety maintained yields under harsh test conditions. They did this by engineering the corn to express a protein from the bacterium Bacillus subtilis, an extra gene that boosted yields by 6 to 10 percent. Monsanto even got permission from the Mexican government to plant test plots in a range of northern states — a significant concession, since GM maize had been banned in Mexico for more than a decade.
Early in 2015, India, formerly a staunch GM opponent, reversed policy under its newish prime minister, Narendra Modi. Genetically modified mustard was planted at the Indian Agricultural Research Institute, and was expected to be released for sale by mid-year. “Field trials are already on because our mandate is to find out a scientific review, a scientific evaluation,” Environment Minister Prakash Javadekar told Reuters late in February. “Confined, safe field trials are on. It’s a long process to find out whether it is fully safe or not.” Allowing GM crops is critical to Modi’s goal of boosting dismal farm productivity in India, where urbanization is devouring arable land. When Modi was chief minister of Gujurat State, he had already allowed GM cotton to be planted there, and it was, according to state officials, a smashing success.16
Sound policy should separate the two streams of genetic modification, first by refusing to grant patents for any attempts to create varieties that are merely resistant to one or other proprietary pesticide, or that confer inbuilt commercial advantage, or that seek to control life forms. However, we should also recognize that genetic manipulation can be beneficial, and approve attempts to create transgenic varieties that increase yields, confer resistance to viral diseases, and thrive in poor farming conditions — and from which farmers can derive their own seeds.
Make Irrigation-System Responsibility a Local Affair
The last way of improving water productivity is a simple management change: move the responsibility for managing irrigation systems from government agencies to local water users’ associations. This suggestion, from the head of the Earth Policy Institute, Lester Brown, has been amply proven in many jurisdictions:
Since local people have an economic stake in good water management, they typically do a better job than a distant government agency. In some countries, membership includes representatives of municipal governments and other users in addition to farmers. Mexico is a leader in this movement. As of 2002, more than 80 percent of Mexico’s publicly irrigated land was managed by farmers’ associations. One advantage of this shift for the government is that the cost of maintaining the irrigation system is assumed locally, reducing the drain on the treasury. This also means that associations need to charge more for irrigation water. Even so, for farmers the advantages of managing their water supply more than outweigh this additional expenditure.17
In sharp contrast to ecologist Garrett Hardin’s tragedy of the commons theory, when local farmers assume responsibility for the water table, they generally don’t rush to deplete it. Instead, they tend to conserve it, using less water overall.
People worry about agriculture more than any other aspect of the water crisis. How will we find the water to grow the food we all need? After all, droughts have killed millions for centuries — and there are more of us all the time, and more droughts too. But with careful management and sufficient resources, the food problem is tractable. Even in arid parts of the world.